Recent trends of smart nonintrusive load monitoring in buildings: A review, open challenges, and future directions

Abstract

Smart nonintrusive load monitoring (NILM) represents a cost-efficient technology for observing power usage in buildings. It tackles several challenges in transitioning into a more effective, sustainable, and digital energy efficiency environment. This paper presents a comprehensive review of recent trends in the NILM field, in which we propose a multiperspective classification of existing smart NILM techniques. More attention is devoted to describing the contributions of deep learning, feature extraction, computing platforms, and application scenarios for NILM development. Accordingly, NILM technical aspects are first Int J Intell Syst. 2022;37:7124-7179. wileyonlinelibrary.com/journal/int investigated, including data collection devices and public data sets. Next, event-based and non-eventbased NILM algorithms are overviewed. Furthermore, potential limitations of existing solutions are identified, highlighting their technical challenges, especially those related to security and privacy preservation, data scarcity, results reproduction, and implementation and business difficulties. Lastly, future directions are explored to overcome the identified limitations.

A Smart Home Energy Management System Utilizing Neurocomputing-Based Time-Series Load Modeling and Forecasting Facilitated by Energy Decomposition for Smart Home Automation

“…Refs. [10], [20] present a comprehensive review of recent trends in energy decomposition. Smart meter data gathered in, for instance, residential environments and analyzed through energy decomposition based on signal processing and AI techniques can support several useful user-centric use cases [10], [20], [21].…”

“…[10], [20] present a comprehensive review of recent trends in energy decomposition. Smart meter data gathered in, for instance, residential environments and analyzed through energy decomposition based on signal processing and AI techniques can support several useful user-centric use cases [10], [20], [21]. One of these use cases is home automation, including anomaly detection for home security [22] and the recognition of activities of daily living (ADLs) for healthcare applications [22], [23]- [25].…”

“…One of these use cases is home automation, including anomaly detection for home security [22] and the recognition of activities of daily living (ADLs) for healthcare applications [22], [23]- [25]. The current techniques adopted in HEMSs for such home automation applications depend on the installation and use of sensors and/or other devices such as video cameras, which also require additional installation and annual maintenance costs [20]. However, an HEMS that exploits energy decomposition for home automation purposes (based on the premise that electrical energy consumption patterns are indicative of residents' daily lives) does not require such additional costs.…”

Thresholding methods in non-intrusive load monitoring

“…1 Soon after these datasets became available, the prevailing approach to NILM shifted from a combinatorial optimization approach to a supervised learning problem with times series in machine learning (ML) [10,11]. Traditional ML methods such as hidden Markov models were initially used [12][13][14], while in recent years the incredible growth of deep learning (DL) algorithms has dominated the field [15]. Some popular architectures applied to NILM are recurrent neural networks, which are designed to process sequential data [16][17][18][19].…”

Meta-learning-based few-shot identification for novel loads

“…It senses power consumption appliances through electrical data such as current and voltage. Based on the results of load identification, it can provide users with energy-saving power consumption suggestions, identify the operating conditions of appliances to monitor the health status, and provide a foundation for grid operators to develop demand response strategies [2].…”